This study was conducted using machine learning models to identify patient non-invasive information for cardiovascular complications prediction in peritoneal dialysis patients. Nowadays is well known that cardiovascular diseases are the key to mortality in patients undergoing peritoneal dialysis as the risk of cardiovascular disease increases with the progression of renal failure. Primary aim is to establish variables most associated with cardiovascular complications. To achieve this goal four different machine learning techniques were used. We found that the best classification algorithm was a Generalized Linear Model, which achieved AUC values above 96% using a small subset of the original variables following a feature selection approach. Our approach allows us to increase the interpretability of the combinations of traditional factors, advanced chronic kidney disease factors and peritoneal dialysis factors all related with cardiovascular risk profile. The final model is based primarily in the traditional factors.

Percutaneous vascular access can be used temporarily (temporary central venous catheters) or permanently (tunneled central venous catheters) [1]. Although percutaneous vascular access is considered inferior to native vascular access due to its shorter half-life and its more frequent complications, the use of percutaneous accesses has increased signifi cantly in recent years. This is due to the fact that the renal patient has more cardiovascular complications, more morbidities and more advanced age, in addition to the waiting time until the AVF maturation which in some patients needs more time due to associated comorbidities such as diabetes mellitus, peripheral arteriopathy, smoking, obesity, advanced age and suboptimal vascular anatomy; even reaching the primary failure of the native access created. In addition to this the easy access to its placement and the immediacy of its use for hemodialysis, has allowed the abuse of its use as vascular access [2,3]. According to the published results of some studies, the use of central venous catheters has increased in many countries, with the percentage of patients dialyzing through tunneled catheters as high as 27.7% in Sweden, 35% in Belgium and 49.1% in Canada [3,4]. The ideal site for the placement of permanent catheters is the right jugular vein, but in some cases it is impossible to

Adapted automated peritoneal dialysis (aAPD), comprising a sequence of dwells with different durations and fill volumes, has been shown to enhance both ultrafiltration and solute clearance compared to standard peritoneal dialysis with constant time and volume dwells. The aim of this non-interventional study was to describe the different prescription patterns used in aAPD in clinical practice and to observe outcomes characterizing volume status, dialysis efficiency, and residual renal function over 1 year. Prevalent and incident, adult aAPD patients were recruited during routine clinic visits, and aAPD prescription, volume status, residual renal function and laboratory data were documented at baseline and every quarter thereafter for 1 year. Treatments were prescribed according to the nephrologist’s medical judgement in accordance with each center’s clinical routine. Of 180 recruited patients, 160 were analyzed. 27 different aAPD prescription patterns were identified. 79 patients (49.4%) received 2 small, short dwells followed by 3 long, large dwells. During follow-up, volume status changed only marginally, with visit mean values ranging between 1.59 (95% confidence interval: 1.19; 1.99) and 1.97 (1.33; 2.61) L. Urine output and creatinine clearance decreased significantly, accompanied by reductions in ultrafiltration and Kt/V. 25 patients (15.6%) received a renal transplant and 15 (9.4%) were changed to hemodialysis. Options for individualization offered by aAPD are actually used in practice for optimized treatment. Changes observed in renal function and dialysis efficiency measures reflect the natural course of chronic kidney disease. No safety events were observed during the study period.

No abstract